Introduction

Ancient DNA represents a direct window to study evolutionary processes of the past. The information stored in these ancient biomolecules can provide insights into, for example, migrations and colonisation events of early humans (Haak et al. 2010; Rasmussen et al. 2011), phylogenetic relationships of extinct species (Shapiro et al. 2002; Bunce et al. 2009), the taxonomic composition of ancient ecosystems (Willerslev et al. 2003; Hebsgaard et al. 2009), or uncover individual phenotypic traits like eye colour, skin colour or diseases (Rasmussen et al. 2010; Keller et al. 2012), which are not preserved in ancient remains. Numerous studies in recent years have documented the prolific scientific synergisms occurring in the overlap between molecular biology, archaeology and palaeontology.

Working with ancient and degraded DNA is not without challenges. Most serious perhaps is the risk of contamination with contemporary DNA, which can easily 'swamp' the small amount of authentic DNA in an ancient sample. This fact became painfully clear in the earlier years of ancient DNA research, when a number of high profile publications seemingly presented evidence of DNA from samples of extreme age—including 80-million-year-old dinosaur bones (Woodward et al. 1994) and >120-million-year-old insects (Cano et al. 1993). We are now aware that these results reflected modern contaminants (reviewed in Willerslev & Cooper 2005), in part because the rate of post mortem fragmentation of DNA excludes the existence of DNA that old (Allentoft et al. 2012). In response to these claims, a set of ancient DNA 'rules' was established to minimise the risk of contamination and to provide some means for downstream authentication of the results (Cooper & Poinar 2000; Gilbert et al. 2005a) and, since then, numerous studies have identified and assessed contamination problems in ancient samples (e.g. Gilbert et al. 2005b; Malmström et al. 2005; Yang & Watt 2005; Pilli et al. 2013).

In essence, however, most of these rules represent 'symptomatic treatment'. If a sample is contaminated with modern DNA before entering the laboratory, for example during excavation, it can be difficult to remove the contamination, and it can be impossible to distinguish between authentic ancient DNA and contaminants. This latter problem is particularly pertinent if the target DNA and the contaminant DNA are from the same species, as is often the case when research is conducted on ancient human material.

In the era of high-throughput sequencing, statistical tools based on DNA damage signatures are now available to assess the overall authenticity of 'bulk' ancient DNA data (Ginolhac et al. 2011). Considerable time and resources, however, are expended before the contamination can be identified, and if a sample has been contaminated decades or centuries ago it may result in contaminant DNA that displays degradation damage patterns similar to that of true ancient DNA (short fragments and an increase in C to T or G to A base changes close to the end of the sequences) (Sampietro et al. 2006; Garcia-Garcera et al. 2011). Although strict laboratory exercises are often combined with bioinformatic analyses to respectively remove and identify DNA contamination, it would be extremely beneficial if the risk of contamination could be lowered in the first place.

In addition to contamination risks it is relevant to consider the negative impacts on DNA preservation that can happen in a specimen after excavation. It has been demonstrated that DNA in bone, to a greater (Pruvost et al. 2007) or lesser extent (Allentoft et al. 2012), will degrade faster when it is removed from its preservation environment. Even the most powerful sequencing technology is useless when the DNA strand is completely disintegrated, so an excavated specimen must be treated carefully to minimise this expected increase in molecular decay.

Below are listed five simple rules that can be implemented during an archaeological excavation to increase the potential for downstream ancient DNA analyses. They are not novel per se, since somewhat similar guidelines can be found in previous articles, or as part of more comprehensive ancient DNA guidelines (e.g. Reference BrownBrown 1998; Yang & Watt 2005; Brown & Reference Brown and BrownBrown 2011; Matisoo-Smith & Horsburgh 2012). Arguably, however, these are the five most crucial rules that should be applied to maintain the molecular integrity of ancient samples. Equally important for the field archaeologist, these five rules are easy to implement by default at any given excavation (DNA or non-DNA focused) with a minimum of preparation and delay. I note that the five rules represent minimum requirements but there are other aspects to consider. If preparing for an excavation that is specifically aimed at recovering material for ancient DNA analyses, it would be beneficial to consult more detailed literature on this subject.

The five rules

1. Minimise the time between initial exposure (discovery) and removal of the specimen from the sediment. If, for recording purposes, the specimen must remain a little while in situ, it can be covered with plastic until removed.

   Reason: the DNA contamination risk increases when a specimen is exposed.

2. Handle the specimen with plastic gloves, and wear disposable sleeves and a face mask. If any tools are used to extract/remove the specimen from the sediment these should be sterilized in 10 per cent hypocholorite before use.

   Reason: these procedures will minimise the risk of contamination from modern human DNA.

3. Place the specimen in a zip-lock plastic bag or other sterile container. If it is subsequently necessary to take out the sample for further examination, wear gloves, disposable sleeves and a face mask.

   Reason: as above.

4. Keep the specimen cold and dry (preferably frozen) until the molecular work can be carried out at a dedicated ancient DNA facility. Never wash the specimen and/or leave it to dry in the sun.

   Reason: heat and humidity will increase the rate of pure chemical DNA fragmentation (hydrolysis and oxidation) as well as microbially facilitated DNA degradation (enzymatic).

5. Minimise the time period from excavation to DNA extraction.

   Reason: in general, the DNA degrades faster in a sample no longer in situ.

Concluding remarks

A trained ancient DNA researcher will by default assume that any given sample is contaminated with modern DNA and will therefore treat it accordingly. An implementation of the above guidelines should not change that practice. Nonetheless, if more caution is exercised in both the field during excavation, and later during cataloguing and storage, the proportion of recovered specimens that display a molecular scientific value should increase significantly.